MATRIX_1.0 live cells record real trials; historical 0.13 scores are not copied.
DynosAI 1.0.0-rc.10 publishes MATRIX_1.0 with four provider-aware cells. Codex greenfield and brownfield are required for 1.0. Cursor cells stay as Preview evidence and are not 1.0 certified. None of them copy 0.13 Quality 100 scores. A cell is pass only after a real trial. The historical table further down is the 0.13.0 core baseline, not 1.0 live proof.
1.0 eligibility vs Preview evidence
| Provider | Integration | 1.0 role |
|---|---|---|
| Codex | App Server + MCP | Certification required (greenfield and brownfield) |
| Cursor | CLI/ACP + MCP | Preview / non-blocking; not 1.0 certified |
MATRIX_1.0 live certification
| Provider | Mode | Status |
|---|---|---|
| codex | greenfield | pass |
| codex | brownfield | fail |
| cursor | greenfield | fail |
| cursor | brownfield | fail |
Historical 0.13.0 core matrix
| Provider | Scenario | Result | Quality | Oracle |
|---|---|---|---|---|
| Codex | Greenfield · Fibonacci | PASS | 100 | 8/8 |
| Cursor | Greenfield · Fibonacci | PASS | 100 | 8/8 |
| Codex | Brownfield · Contract discounts | PASS | 100 | 10/10 |
| Cursor | Brownfield · Contract discounts | PASS | 100 | 10/10 |
What the matrix proves
The historical 0.13.0 matrix proved both supported providers could complete governed greenfield work with the then-current workflow.
Both preserved and extended existing behavior in the brownfield scenario while passing independent Oracle checks.
Codex used structured-primary MCP transport while Cursor used the compatibility transport required by its CLI stream. That is not 1.0 live proof.
What it does not prove
It is not a universal guarantee across every language, repository size, CI system, security policy, or provider version.
DynosAI 1.0.0-rc.10 records MATRIX_1.0 from real trials or leaves cells honestly not_run. Historical 0.13 evidence is not copied. It is not a production-ready 1.0 claim.
The Predictive Router remains in shadow mode even though the historical sample size now crosses its quantitative authority gates.